基于地面高光谱遥感与数字图像信息融合的甜菜氮素诊断方法研究

项目名称： 基于地面高光谱遥感与数字图像信息融合的甜菜氮素诊断方法研究

项目编号： No.41261084

项目类型： 地区科学基金项目

立项/批准年度： 2013

项目学科： 天文学、地球科学

项目作者： 田海清

作者单位： 内蒙古农业大学

项目金额： 48万元

中文摘要： 常规的化学测试很难满足丰产高糖条件下甜菜氮素营养实时诊断的要求。地面光谱遥感具有实时、无损诊断植株氮素的优势，但甜菜叶丛形成期前植被盖度低，光谱易受土壤等背景信息干扰，而块根及糖分增长期后易出现光谱植被指数"饱和"问题，影响诊断精度。在其它作物氮素诊断中这些影响因素同样存在，目前尚无理想的解决方法。甜菜冠层较低，叶片肥大，上述问题尤为突出。考虑到高光谱分辨率高、信息量大的优点，而数字图像便于背景分割、颜色特征与作物氮素水平明显相关的特点，本项目借助多信息融合比单一传感器更优越的性能，结合植株氮素测试，开展基于高光谱遥感与数字图像信息融合的甜菜氮素诊断方法研究。在深入研究氮素敏感光谱特征参数提取、图像背景分割及氮素敏感颜色特征参数提取方法的基础上，融合高光谱与图像信息，校正土壤等背景干扰信息影响，解决传统光谱指数易"饱和"的问题，建立甜菜氮素实时诊断体系，为甜菜氮素的实时诊断提供新方法。

中文关键词： 甜菜；氮素诊断；高光谱遥感；图像；信息融合

英文摘要： Traditional lab-based chemical testing is difficult to real-time estimate N status of sugar beet under high yield and high sugar conditions. Ground-based spectral remote sensing is characterized with real-time, non-destructive diagnosis of plant nitrogen status. However, the spectra are disturbed by soil background before foliage formation stage, and spectral indices easily lose sensitivity after tuberous roots and sugar growth stage of sugar beet. These factors also exist in other crop nitrogen nutrient diagnosis. Due to the lower canopy and leaves hypertrophy for sugar beet, the problem is particularly prominent. Taking into account the high spectral resolution, the advantages of a large amount of information and digital images to facilitate the segmentation of the background color characteristics, we perform multi-information fusion based N nutrient diagnosis in present study. In-depth study of nitrogen-sensitive spectral features variable extraction, image background segmentation and feature parameter extraction, we attempt to decrease the disturbance of soil background and solve the problem of saturation for the traditional spectral index. The final aim is to establish a rapid diagnostic system of nitrogen for sugar beet production, and to provide a new method for real-time diagnosis of beet nitrogen.

英文关键词： Sugar beet；Nitrogen nutrient diagnosis；Hyperspectral remote sensing；Image；Information fusion